OpenAI details a separate ChatGPT ads stack built on a post-trained LLM
Original title: How we built ChatGPT’s novel ad stack for user trust and relevance
OpenAI says ChatGPT Ads reached a $1 billion annualized run rate in under 200 days, with over 100,000 advertisers.
OpenAI published an engineering account of ChatGPT’s advertising system. A stateful session and a post-trained LLM orchestrate retrieval, ranking, relevance checks, and auctions through tool calls. Ad selection runs in a separate service so sponsored placements do not shape answers, and an unavailable ads service does not block ChatGPT responses. OpenAI says returning no ad is a normal outcome.
Unconfirmed: The revenue and advertiser totals are OpenAI’s own claims, not independently verified.
- OpenAI says more than 100,000 advertisers have run ChatGPT ads since launch.
- The company claims a $1 billion annualized revenue run rate in under 200 days.
- Ads sessions retain context across turns and across conversations.
- ChatGPT and ads use separate services, models, and code.
From the original article
13 min read at openai.com
Learn how OpenAI built ChatGPT’s LLM-powered ad stack to understand conversational intent, keep ads separate from answers, and deliver relevant ads at scale.
openai.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.